One moving image encoding device that exists is a moving image inversion encoding device that performs encoding after symmetrically inverting an input image in a vertical direction, a horizontal direction, or a vertical and horizontal direction frame by frame beforehand as mentioned in Non Patent Document 1. This is because input images generally have non-uniform characteristics within the frames and, therefore, in cases where predictive encoding with directional dependence from the top left to the bottom right is performed in raster scan order, the encoding efficiency can be improved by inverting the image in the direction of higher predictive efficiency.
The encoding device according to Non Patent Document 1 inverts the input image by choosing from four types of options, namely, no inversion, vertical inversion, horizontal inversion, and vertical and horizontal inversion. (a) of FIG. 1 shows the input image; (b) of FIG. 1 shows an image resulting from the non-inversion of the input image; (c) of FIG. 1 shows an image obtained by inverting the input image in a vertical direction; (d) of FIG. 1 shows an image obtained by inverting the input image in a horizontal direction; and (e) of FIG. 1 shows an image obtained by inverting the input image in a vertical and horizontal direction. Furthermore, in inter-predictive encoding which makes predictions between frames, the encoding device inverts the reference image which is a locally decoded image of an image which has been previously encoded in the same direction as the input image and performs encoding by using the inverted reference image to generate a predictive signal for the inverted input image constituting an encoding target. Compressed data obtained by encoding are output together with an identifier which indicates the inversion direction of the input image. In addition, the compressed data obtained by encoding are locally decoded, inverted in the opposite direction from the inversion direction of the input image, translated in the same direction as the input image, and saved as a reference image for the next frame and subsequent frames. (f) of FIG. 1 shows an image obtained by inverting the input image ((b) of FIG. 1) resulting from non-inversion of the input image in the opposite direction. That is, (f) and (b) of FIG. 1 show the same image as (a) of FIG. 1. (g) of FIG. 1 shows an image which is obtained by inverting an image ((c) of FIG. 1) obtained by inverting the input image in a vertical direction in the opposite direction. (h) of FIG. 1 shows an image which is obtained by inverting an image ((d) of FIG. 1) obtained by inverting the input image in a horizontal direction. (i) of FIG. 1 shows an image which is obtained by inverting an image ((e) of FIG. 1) obtained by inverting the input image in a vertical and horizontal direction in the opposite direction.
Furthermore, in inter-predictive decoding which makes predictions between frames, the decoding device which appears in Non Patent Document 1 discriminates the inversion direction by obtaining an identifier that indicates the inversion direction of the input image from code which is transmitted by the encoding device and inverts a reference image which is a previously decoded image in the same direction as the inversion direction of the input image. Further, the decoding device uses the compressed data obtained from the code transmitted by the encoding device and the inverted reference image to decode the inverted image and then inverts the inverted image in the opposite direction from the inversion direction of the input image to translate the image in the same direction as the input image and generate a decoded image, and saves the decoded image as a reference image for the next frame and subsequent frames.
[Non Patent Document 1] Murakami et al, “Expanded encoding system that utilizes directional dependence of a H.264/AVC Baseline Profile”, IEICE 2005